@isanuage

Achieving Agility and Scale for Your Data Lake

Isabelle Nuage, Product Marketing Cyril Sonnefraud, Product Management

©2017 Talend Inc #TalendConnect

Poll• Who’s using Talend Big Data today?• Who has a data lake in production?• Who is deploying or planning a data lake project within

12 months?• Who is implementing a data lake in the Cloud?

<Digital Tranformation Stats>

By end 2017, > 70% of G500

By 2020, 50% of the G2000

Digital Transformation is no Longer an Option Are You Prepared?

But only 26% of Organizations

Accenture and Forrester Digital Transformation in the Age of the Customer studyIDC Futurescape

The Data Lake is the New Digital Backbone

• Break down data silos• Structured and

unstructured • Granular data• Machine learning

Bu

sin

es

s V

alu

e• Offload EDW

• Cheaper storage

• Access to archived data

Why Create Data Lakes?

Reduce costs

Bu

sin

es

s V

alu

e

Generating new opportunities

• Offload EDW

• Cheaper storage

• Access to archived data

• Customer acquisition, retention..

• Real-time engagement

• Pricing optimization

• Demand forecasting

• Risk and fraud

• Predictive maintenance

• Smart products…

Why Create Data Lakes?

Reduce costs

7

Challenges

Complex Technology

Limited Access

DataSwamps

How to achieveAgility & Scale?

DATALAKES

#TalendConnect

People Doing it the OLD Way…

#TalendConnect

2017 Lenovo Internal. All rights reserved.

Change is the Only ConstantB

usin

ess Valu

e

Reporting MeasurementBusinessInsights

Optimization Predictive Analytics

Automation Prescriptive Analytics

Pre FY - 07FY - 07/10

FY - 11/ 12

FY - 13/ 14

FY – 15/ 17

Time

Cognitive Analytics

FY – 17/ 18

• Any innovation

• Any platform

• Any use case

• Any speed

• Any user

The Agile Data Lake

The Path to Agility

Ing

esti

on

+

basic

vis

uali

za

tio

n

Data

Qu

ali

ty

Self

Serv

ice

Data

Go

vern

an

ce

Real-

tim

e

Mach

ine

Le

arn

ing

©2017 Talend Inc #TalendConnect

ExamplesSmart Data QualitySmart Data Pipelines

Demo flow

Data Lake

Incoming Lead Data

(Raw)

Amazon EMR Cluster Data Lake

Output Lead Data (Processed)

With Segmentation

1

Ingestion with Smart Data

Quality

2

Smart Data Pipeline

with Machine Learning

©2017 Talend Inc #TalendConnect

ArchitectureGuidelines

On-premise Data Lakes

On-Premise

Data Sources

Ingest Prepare Process Access Consume

Cloud

Data Sources

Governance

Processing

Storage

On-prem Datalake

Hybrid Data Lakes

On-Premise

Data Sources

Ingest Prepare Process Access Consume

Cloud

Data Sources

Governance

Cloud Processing

Processing

Cloud Storage

Storage

On-prem Datalake

Cloud Datalake

Distribute

Cloud Data Lakes – A Concrete Example

Ingest Prepare Process Access Consume

Governance

Cloud Processing

Cloud Storage

On-Premise

Data Sources

Cloud

Data Sources

S3

EMR

Cloud Storage

Cloud Dataflow

Azure DL Store

HDInsight

The Path to Agility

Ing

esti

on

+

basic

vis

uali

za

tio

n

Data

Qu

ali

ty

Self

Serv

ice

Data

Go

vern

an

ce

Real-

tim

e

Mach

ine

Le

arn

ing

Deliver Value Along The Way

Start with quick wins & business outcome in mind

Get a cadence of constantly delivering value

Focus on game changer value drivers

Get the company onboard

Be Eligible to Win Prizes at the End of the Show!